Error: Function value and YDATA sizes are not equal

I have a fitting code with lsqcurvefit function. I am having this error "Function value and YDATA sizes are not equal"(line 251). Please give your suggestion to solve it.

clear all, close all,  clc
EQE = xlsread('EQERollOff','Sheet1','A30:B55');% loading raw data from excel file 
x = EQE(:,1);% Current
y = EQE(:,2);% EQE
 %parameter estimation
J0 = [6]; 
options = optimset('Algorithm', 'trust-region-reflective');   
[J,Rsdnrm,Rsd,ExFlg,OptmInfo,Lmda,Jmat] = lsqcurvefit(@EQEFit,J0,x,y)

function s = EQEFit(J,x)% function for fitting
s = J(1)/(4*x)*(sqrt(1+8*(x/J(1)))-1)*5
end

Size of x = 26 1, Size of y = 26 1

Answer 1

[Rik]

As a bold guess I'm going to assume you meant this:

function s = EQEFit(J,x)% function for fitting
s = J(1)./(4*x).*(sqrt(1+8*(x/J(1)))-1)*5;
end

(I don't have the Optimization Toolbox, so I can't test the code)

[Monirul Hasan]

Hi Rik, it worked. Thanks for your help. Just wondering as J(1) is scalar then why J(1)./(4*x) did work? I am new in in coding. Please explain so I can avoid doing the same mistake again. Cheers

[Rik]

There are two operations here that are potentially matrix operations: the division and multiplication. Since the multiplication returns a vector, this would be equivalent to finding an inverse for that vector and then multiplying that with J(1).

so this:
J(1)/(4*x)
is parsed like this:
J(1) * 1/(4*x)

The most important thing is to be aware that these operators have two imputs: the prior term and the second term.

The more practical thing is to watch out with these, and adding a dot when in doubt:

*/^'

[Monirul Hasan]

Thanks Rik.